ML17297B533
| ML17297B533 | |
| Person / Time | |
|---|---|
| Site: | Palo Verde  |
| Issue date: | 05/10/1982 |
| From: | Licitra E Office of Nuclear Reactor Regulation |
| To: | Office of Nuclear Reactor Regulation |
| References | |
| NUDOCS 8205200357 | |
| Download: ML17297B533 (46) | |
Text
Docket Nos: 50-528/50-529 and 50-530 WY ao
~982
".<c
+@~ lg)g
.0 e~r ~~2m.
p 4~y@ejf Aye APPLICANT:
Arizona Public Service Company Co FACILITY:
Palo Verde Nuclear Generating Station Units 1, 2
SUBJECT:
SUMMARY
OF APPEALS MEETINGS HELD ON DECEMBER 8, 1981 On December 8, 1981 the staff met with the applicant to discuss the following issues:
1.
appeal'f staff position on tornado missile protection for UHS spray nozzels; 2.
appeal of staff position on DG fuel oil tank corrosion protection; 3.
appeal of. staff position on UHS capacity.
A list of attendees is inclused as Attachment 1.
Tornado Missile Protection The applicant discussed the probability of a damaging tornado stri-ke to the UHS sp jay nozzels...-
Attachment 2 presents the applicant's discussion.
The applicant agreed to formally submit the information contained in Attachment 2.
The formal submittal must contain statistics corrected for population density, must discuss whether two nozzle sets can be impacted by a missi'te, and must demonstrate retention of sufficient heat removal capacity in the worst case.
A final staff decision will be made after staff review of the formal submittal.
DG Fuel Oil Tank Corrosion The applicant was concerned that a fuel oil tank coating to inhibit corrosion would flake off.
The staff felt that this was not a significant problem.
However, the staff agreed to consider any evidence that the applicant could submit which would demonst;rate that the humidity in Arizona is sufficiently low so that DG fuel tank corrosion is not a problem.
The staff suggested that experience with similiar type, tanks would be sufficient:,- If objective data could not be provided by APS, then the staff would require that the DG fuel oil tanks be coated.
OFFICE/
SURNAME/
DATEP 820520
< p5ppp528
-p387 82p5 PDP P,DQGK pop A
~
~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~
~
~
~ ~ ~ ~ 0
~ ~
~ ~ ~ ~ ~ ~
~ WOO
~ ~ ~ ~ ~ ~
~
~
~ ~ ~ ~
~ ~ ~ ~ ~
NRC FORM 3)8 l)0/80) NRCM 0240 OFFICIAL RECOFiD COPY e USGPO: )980-329.824
I'4>>
J
$ j I //
44
/
/4
~/I
~ I
~ 4 4
J I
4
~ it /
~,'"
ll C I" JJJ.
4
~I 4 4
~ j ~
a/
~
= ~
~I I
4
~
4 JJ I
'I I I
ghJJJ 4
t/4
'I 4 444 I/
~
I
/
J ')I
~
r.
4="
~
'I ~
I 4
Ia=
J 4
J I
~
4~
'I H
4 4
)
4
~
J /i I
/
P
~
/
I
/ ~
I
-4 lrl 4
/ ~
4
q,% UHS Ca acit APS calculates a seismic water supply sufficient for 26-28 days of operation.
The staff agreed that a 26-28 day water supply in the UHS pond is acceptable provided that:
a) the staff review concurs with this calculation; b)
APS identifies other sources of water that can be used after the spray pond is depleted, including a discussion of the effects of the initiating event on the availability of those sources of water; and c)
APS establishes operating and maintenance procedures that provide assurance that these additional sources of water can be used in the event they are needed.
APS agreed to document the above information.
Attachments:
As stated Ox.'~QeL Sfgue43gg I
E. A. Licitra, Prospect Manager Licensing Branch No.
3 Division of Licensing cc w/attachments:
See next page See attached distribution.
OFFICE)
SURNAME)
DATE)
DL:LB(
MLicit a.p 5//g /82 FMi 5/
l.a.....
/82 NRC FORM 318 IIO/80) NRCM 0240
-OFFICIAL RECORD COPY
- USGPO: I980-329.824 s
>Yj!< 9
~ %Ah!
~
~
4 I
Hr. E.
E.
Van Brunt, Jr.
Vice President - t/uclear Projects Arizona Public Service Company P. 0.
Box 21666 Phoenix, Arizona 85036 PALO VERDE CC:
Arthur C. Gehr, Esq.
Snell
& Wilmer 3100 Valley Center Phoenix, Arizona 85073 Charles S. Pierson Assistant Attorney General 200 State Capitol 1700 West Washington Phoenix, Arizona 85007 Charles R. Kocher, Esq., Assistant Counsel James A. Beoletto, Esq.
Southern California Edison Company P. 0.
Box 800
- Rosemead, California 91770 margaret Walker Deputy Director of Energy Programs Eco'nomic Planning and Development Office 1700 West Washington Phoenix, Arizona 85007 Hr. Rand L. Greenfield Assistant Attorney General Bataan Hemorial Building Santa Fe, t,'ew Mexico 87503 Resident Inspector Palo Verde/NPS U.S. Nuclear Regulatory Commission P. 0.
Box 21324 Phoenix, Arizona 85001 tIs. Patricia Lee Hourihan 6413 S. 26th Street
- Phoenix, Arizona 85040
P
~ ~
'l
'I L~
J. Kerrigan F. Miraglia J. Allen W.
Bengham E.
E.
Van Brunt, Jr.
W. Quinn J.
Vorees Olan Parr L. Rubenstein
.G.. Kopochirmki..
I. Spiekler J. Costello E. Markee J.
Wermiel V. Panciera J.
Koch Julius Goodman Dennis Keith Paul Barbour R. Tedesco M. Fliegel R. Gonzales Richard Codell J.
Knight ATTENDEES
. Attachment.I AFFILIATION NRC NRC APS Bechtel-APS APS APS NRC NRC Bechtel NRC
'NRC NRC NRC NRC Bechtel Bechtel Bechtel Bechtel NRC NRC NRC NRC NRC/DE
r
~
QCAFt Attachment 2
PALO VERDE NUCLEAR STATION TORNADO HA'ZARD TO STATION ULTIMATE HEAT SINK 1.
Task Description SER Section 9.2.5 open item states "..
. The.applicant has not provided sufficient information to alleviate the staff concern on the lack of tornado missile protection for the essential spray pond spray nozzles.
It is the staff position that this applicant demonstrate the capability of the'HS to 'safely 'p'erformts safe shutdown function in"the 'event t
k
~
~
of' tornado The spray ponds and spray nozzle configuration are depicted in figures 1-5.
2.
Acceptance Criteria GDC 2
{ 1) requires "Structures,
- systems, and components important to safety shal.l be designed to wi thstand the effects of natural phenaaena such as - tornados
- without loss of capability to perform their safety functions SRP 3.5. 1.4 (2) lists an acceptance probability for tornado damage of
-7 10 below which pr otective structures are not required.
3.
Summary and Conclusions The analysis considers mi The damage probabilities ssile damage to various numbers of spray nozzles.
are presented in table I.
For conservatism, the upper limit (95$ ) is used for ccmparison with the acceptance criteria
-7
-7 of 1 x 10
.- The upper limit damage probability of 1 x 10 per year corresponds to damage of in one spray pond train.
about 6$ of the total number of nozzles sets A separate analysis determined that this would have a very minor effect on the flow through the intact nozzles.
Therefore, approximately 94$ of the heat 'rejection capability of the pond is available.
~
~
4.
Assumptions and Conservatisms The, following assunptions were used in this study:
a.
Tornado characteristics are in accordance with the values for Region II cited in Regulatory Guide 1.76 (3).
b..The missiles of Classes A, B, C, D, E and F according to Standard Review Plan (2) are considered.
c.
Class 6 missiles (auto) are excluded from consideration because the parking area is far from the spray pond and the toggado of the
>~> C98LES ACET p
~ pF credible intqnsity cannot..transport them such distance.
d.
The distributions of potential missiles by numbers and lengths are adopted from EPRI Study (4).
Conservatisms incorporated in this study:
a.
Damage probabilities are reported on one of two fully redundant spray trains.
b.
Conditional probability of damage to the nozzles given hit is equal to one.
c.
The effective areas of targets are overestimated several times.
d.
The distribution of potential missiles is scaled on maximum value
.according to survey data.
e.
l'he possibility of creation of missile cluster with density 10 times exceeding the upper limit is considered.
f.
A missile strike in the effective target area damages all four I
nozzles of a nozzle set.
5.
Analysi s Approach The analysis took into account the uncertainty associated with each factor by use of probability distribution curves for each.
The median values and 905 confidence intervals are reported.
The probability, P, of damage by a tornado missile impact can be written:
T 0
H D
Where P
is the probability of a tornado occurrence at the plant site, P
is the conditional probability of hitting a target given the tornado occurrence and P
is the condi.tional..probabil.i.ty..of. target. damage assuming a hit.
The method of calculating the tornado occurrence rate P
was proposed by Thorn (5) and recommended by Standard Review Plan (2):
ea 0
S Where:
= annual frequency of tornado occurrence in area S.
a
= tornado path area (sq. mi.)
S
= one degree region including plant site,
~ 4000 sq..mi.
A 31 year record of tornado strikes (6) in the state of Arizona and speci-"
fically the county of Haricopa is used to establish the annual frequency of tornado occurrence (0 ), in area S.
This data is presented in table II.
This data is consistant with generic region II data reported by Thorn (5)
The 31 year record of tornado strikes in the state of Arizona didn' include sufficient data to calculate accurate values of tornado path area (a),
For this reason, the generic path areas determined by Thorn (5) are used in this study.
These path areas are given in table II.
I
4
<<M <<&<<k %<<&.8 ~
~ <<W
~ , 4<<<<
5.
Analysis Approach The analysis took into account the uncertainty associated with each factor by use of probability distribution curves for each.
The median values and 905 confidence intervals are reported.
The probability, P
, of damage by a tornado missile impact can be written:
T 0
H D-Where P
is the probability of a tornado occurrence at the plant site, P
is the conditional probabi1ity of hitting a target given the tornado occurrence,and P
. is. the.'conditional probability of.. target damage assuming a hit.
The method of calculating the tornado occurrence rate P
was proposed by Thorn (5) and reccmmended by Standard Review Plan (2):
P 9a 0
S Where:
= annual frequency of tornado occurrence in area S.
a
= tornado path area (sq. mi.)
S
= one degree region including plant site,
~ 4000 sq..mi.
A 31 year record of tornado strikes (6) in the state of Arizona and speci-fically the county of Haricopa is used to establish the annual frequency of tornado occurrence (0 ), in area S.
This data is presented in table II.
This data is consistant with generic region II data reported by Thorn (5)
The 31 year record of tornado strikes in the state of Arizona didn' include sufficient data to calculate accurate values of tornado path area (a).
For this reason, the generic path areas determined by Thorn (5) are used in this study.
These path areas are given in table II.
The probability of a tornado occurrence at the plant site is determined from the above information.
The results are shown in table II.
The method of calculating the conditional of probability hitting a given target (P
) was developed by Goodman and Koch (7).
In the current H
analysis we are concerned with multiple targets (nozzle sets).
- Thus, the first step in the analysis is the determination of the probability of hitting one target, then calculate the probability of hitting more I
than one target.
The probability of hitting one nozzle set
{P ), at elevation Z, is determined,by the following relationship:
p' H h g4(<)q(+)4(< t-1 H
P F
Where:
N
= the local density of potential missiles near the spray ponds P
(ft
)
A
= the effective target area of one nozzle set (ft )
2 gF) = the relative frequency of tornado with severity equal F on the Fugi ta scale g(F) = the probability that a potential missile beccmes airborne gc~)= height distribution of airborne missiles The number of potential tornado missiles at nuc'lear plants has been evaluated by Twisdale {4).
The study included on-site surveys at seven plants.
In addition to total number of missiles over the entire site, the distribution of'hese missiles over the site was also evaluated.
The current study uses this data and takes the high local density of potential missiles because it coincides with the location of the spray ponds.
Thi s data i s presented in table I I I, The effeet ive area of a
spray nozzle set is conservatively calculated as circular area enclosing the nozzles set, The diameter of this circle is the sun of the diameter of the nozzle set plus the length of the average tornado missile.
This area is 347 ft.
The distribution of relative frequency of tornado 2
( $ (F)) with severity F.on the Fugita scale (8) is taken fran Wen and Chu (9).
The probability that a potential missile beccmes airborne is taken from Goodman 8 Koch (7).
The height distribution of missiles that"do become airborne is based on an exponential relationship AC~~i described in (7).
This later relationship is in good a+gumeat with missile simulation work reported by Twisdale (4).
The sunation of the last three terms of the above equation over all tornado severities (F) yi,el ds. the.probability..that a: potential.,mi ssil e.becomes airborne and.:
elevated to a height Z.
This value is given in table III.
Using the above data, the probability of hitting one nozzle set (P
) is cal-culated and presented in table III.
The next step in the analysis is the calculation of the conditional probability of multiple missile strikes resulting in damage to more than one nozzle set "in one spray train given the tornado.
This probability is obtained from the modi fied bi ncmi al di str ibution, as fo 1 1 ows:
pH-M PH I
PH F
Where:
total number of nozzle sets in one spray train number of damaged nozzle sets t
N
= combinations of undistinguishable nozzle sets that could be damaged
=
N!
N-H!
l -xP 4l So correction factor for missile clustering p
g
Mhere:
~
S
= effective area of local cluster of airborne missiles 1
= missile or1gin area (the maximum area frcm which the potential missiles can affect the target is in considerat1on)
X
= ratio of-density in the cluster to the mean density in the area So Each spray train has eighty nozzle sets (N) consisting of four nozzle, as depicted in figure 3.
The probability of damaging multiple nozzle sets (N) is strongly dependent on the density of airborne missiles.
In 'this study, 'mi sSil e clustering 'was'ostul ated.
Ni ssil e'lUster
densities up to ten times the upper limit density of potential missile in the local origin area are used.
The correction factor (F) is used to adjust the multiple nozzle set damage probability based on statisti-cally independent strike probabilities (P
) to dependent probabilities (P
) resulting from the cluster model.
The conditional probability of H
multiple target damage per year, given the tornado (P
) is given in table IV.
Combining this data with the probab1lity of a tornado occurrence at the plant s1te'P
) and the conservative assumption that P
= 1, gives the probabil1ty of damage to N nozzle sets in one spray train per year.
This data is presented in table V.
6.
References:
(1) 10 CFR part 10, Appendix A, Criterion 2, "Design Basis for Protection Against Natural Phencmena".
(2)
Standard Review Plan, U.S. Nuclear Regulatory Commission, NUREG-75/087, Revision 1, section 3.5.1.4.
(3)
Nuclear Regulatory Commission, "Design'Basis Tornado for Nuclear Power Plants",
Regulatory Guide 1.76, April 1974.
(4)
Twi sdale L.A., Dunn W. L., Clue J., "Tornado Missile Risk Analysis",
EPRI NP-768, May 1978, Twi sdale L.A., Dunn W.L., Lew S.T., Davis J.L.,
Msu J.C.,
Lee S.T., "Tornado Missile Risk Analysis-Appendixes",
EPRI NP-769, May 1978.
(5)
.Thorn..H,.C.S.,
"Tornado Probabilities",,Monthly..Weather.
Reivew,,
Oct.-Dec.
1963.
I tW (6)
U.S.
Tornado Breakdown by Counties 495&-1980, U.S. Department of
- Commerce, National Oceanic and Atmospheric Administration, National Weather
- Service, National Severe Storms Forecast
(7)
J.
Goodmand and J.E.
Koch, "Conditional Probability of the Tornado Missile Impact Given a Tornado Occurrence",
Proceedings of Inter-national ANS/ENS Topical Meeting on Probabilistic Risk Assessment, Port Chester, New York, September 1981.
{8)
Fujita J.J.,
"Proposed Characterization'of Tornadoes and.Hurricanes by Area and Intensity",
SMRP Reserarch Report No. 91, Feb.
1971.
(9)
Wen Y.K. and Chu S.L., "Tornado Risks and Design Wind Speed",
Journal of the Structural Division, Proceedings ASCE,'o.
99, No..C.T. 12, Dec.
- 1973,
TABLE I PROBABILITY OF DAMAGE TO M NOZZLE SETS IN ONE SPRAY TRAIN PER YEAR NUMBER OF NOZZLE SETS-DAMAGED, M
PERCENT OF NOZZLE SETS DAMAGED UPPER LIMIT PROBABILITY (95TH PERCENTILE)
~
1
~
2 3
)0
- 1. 25 2.50 3.75 5.00 6.25 7.50
.8. 75 10.00
)1.25 12.50
.7x)0 1x)0 2x)0 3x)0 Gx)0 2x)0 5x10 1x)0 4 x 10-."
9x10
C1 I'-
"4
~
A I
jI
TABLE I PROBABILITY OF DAMAGE TO M NOZZLE SETS IN ONE SPRAY TRAIN PER YEAR NUMBER OF NOZZLE SETS DAMAGED, M PERCENT OF NOZZLE SETS DAMAGED UPPER LIMIT PROBABILITY (95TH PERCENTILE) 2 7
10
. 1.25
~ A, 2.50 3.75 5.00 6.25 7.50 8.75 10.00 11.25 12.50 7 x 10
~
> ' '. e.,'
i "r" ~
~." '.
1x10 2x10 3x10 6x10 2 x'10 5x10 1x10 4x10 9 x 10-11
TABLE I I PROBABILITY OF TORNADO OCCURRENCE AT PLANT SITE P =~a 0
S Lower Limit Pedi an Upper Limit Annual Frequency
-2 of tornado Occurrence'.89 x 10 0.28
...in Area S. (.'~. 4000<q. mi.)
- 1. 58 g
I
~
Tornado Pith Area {sq. mi.), a Probability of a tornado Occurrence at Plant Site, per year, P 0 1.18 x 10 5.4 x 10 3.45 x 10 2.4 x 10
- 10. 2 1.1 x 10
y~ or5~*'NA'to&'0Av'o"sike;,tlat 'nfl 's to ~mek~>k 4%v ~mores QA 'sAralerremmf ek4' w~o eAkal&ii@sv Cessakfl ssslA sik i'II~44544A~J mmts 4a Rot i e TABLE III CONDITIONAL PROBABILITY OF HITTING ONE TARGET, GIVEN THE TORNADO p'
N u, 54(i)y(i)V(~,~)
H P
F-lower limit Number of potential missile N
on site (2.5 x 10 ft )
1982 7.9 x 10 Ave rage s urface dens ity of
~ potential missile on site s:r Total surface density of 2.0 x 10 potential missile near the spray
- pond, p (ft
)
medi an 2651 1.1 x 10 2.7 x 10 upper limit 3545 1.4 x 10 3.6 x 10 Effective area of one nozzle set, A (ft )
2 Probability that a
potential missile beccmes airborne and elevated to a height Z, 347.
5.3 x 10 347.
7.1 x 10 347.
9.4 x 10 Condi tional probabi 1 ity of hitting one nozzle set given the tornado, 4.9 x 10 6.6 x 10 9.0 x 10
TABLE IV CONDITIONAL PROBABILITY OF MULTIPLE TARGET DAMAGE PER YEAR PH =
MN PH l-PH F
Number of NozzIe Sets
- Damaged, M
'2
'3 10 Lower Limit 4x10
-.6
"-7x10" lx10 lx10 1 x 10-14 2x10 2 x 10-20 3x10 3 x 10-26 3
10 Median 4-x 10
-3 lx}0 2x}0 2x10 4x10 9x}0 2x10 4x}0 6x}0 3x}0 Upper Limit 4x10 lx10 3x10 6x10 1x10 4x10 1x10 5
4x}0 1x10
-6 3x}0
TABLE V PROBABILITY OF DAMAGE TO M NOZZLE SETS IN ONE SPRAY TRAIN PER YEAR Number of Nozzle Sets
- Damaged, M
2 ~,
.7 10 Lower Limit 1.x 10
,5x10, 9 x 10-"
1x10 2x10 3 x 10-22 4 x 10-25 5x10 5
10 3'x10 Medi an 9x10 2.x 10 4x10 6x10 1x10 2x10
-17 y,5x 10 9 x 10 1x10 7 x 10 Upper Limit I'
x 10 1 x.10...
2x10 3x10 6x10 2x10 5 x 10 1 x 10 4x10 9x10
gr 55
~
>e~ey H
I I
I C
I "C
I J
t
~
tt I
~ "~
C 1+Mfll 4C "S
r ~
X., w'<.si rr.
ENP I'ILTEM 5 'NUNP VAI.VEBOX 08TTRRSLY VALVE PRAY NOZZLN 2" TYP
$i"PIPE
~"N%
MAKEUP LIME BUTTERFlY VALVE MAKEUP LINE 346'l ~
ESP TRAIN"B" FIGURE 2
r
2" IPhAYAhlfRYE rm tTVPl q
~ ~
~ ~~,
r~.
SUPPORT POST NW'
~
i"hNKh iTVPf
~g g ELfff'M g TOPOPIPhAYPONDWALLEL.fN'M LOW WATER LEVELEL.10IV-O"
~ ~ ~
~
~ ~
~
~
~
~
~
~
P
~
~
~
~
~ 0> ~ ~ ~
g KLN'M POND FLOOA EL. 92'-S" KStS StRAY SYSTEM FIGURE 3
~
~
I I~
~
~
~
~
~
~
~~~
~<<
,I'I j
<< ~
I<<
4 ~
I<<
4
~
~0 Ib i)
V!
~
i
~
~
~ ~
.L I<<
SPRAY POND LAYOUT FIGURE 4
v+
~ /
/ 4
/~
g(~
SLIDE 3 CLOSE-UP OF SPRAY HEADER 5 NOZZLES FIGURE 5
0" Ck
MEETING
SUMMARY
DISTRIBUTION CDocument Control-(-50-528/529/530)~
LPDR'DR NSIC TERA LBg3 Files IJLee A.
icitra PM
- Attorney, OELD I&E Regional Admi'nistrator, Region V
NRC Participants JKerrigan FMirag 1 ia OParr LRubenstein ISpiekler JCostello EMarkee JWermiel VPanci:era RTedesco MFliegel RGonzales RCodell JKnight cc:
Service List
P <'4r